Vertical & Social: AI Video Models Compared

The format most AI video is actually made in: 9:16 for Reels, TikTok and Stories, shot the way a phone shoots it. Every other category here is 16:9, so this is where models get judged on framing they rarely train for. Each of the 4 vertical & social prompts below ran once on 40 AI video models, including MiniMax FastH3 V2, P-Video 2 Pro and Wan 3.0. Open a prompt to compare every clip side by side.

Vertical & Social

Kerbside Laces 36 models Latte Art Pour 36 models Sneaker Unboxing 36 models Skincare Selfie Demo 36 models